Nexenta - Recovering From A Locked Auto-Sync Service

If replications via the Auto-Sync services are not working, the following is how our third-party support have told me to fix the issue.

1. Browse to the GUI for the device with the failed services.

2. Go to "Data Management" and "Auto Services" - this is a list of all the created Auto-Sync services.

3. Under "Auto-Sync Services" on the left hand side of the screen, click "Show".

4. This page will likely take a while to load so don't worry! Click the tick box next to "Instance" to select all the services, then click "Abort" to stop anything that thinks it is running. Once this has been done, select all the services again and click "Disable". Once again, both of these actions may take some time to complete.

5. Once the services are disabled, use Putty to connect to the device where the replications are being sent.

6. Type "nmc" and press enter to access the Nexenta Management Console.

7. Type the following commands to restart the services;

svcadm restart dbus

This will take quite some time to complete so leave it for 15 minutes to restart all the services

8. Type "svcs" to check the status of the services. If any are showing as offline or disabled, make a note of the service name and run the following command;

svcadm clear servicename

9. Run the following command to restart the rrsync service;

svcadm restart rrdaemon

10. Check the services again by typing "svcs". Once they are showing as online you can close Putty.

11. Back on the GUI for the device with the failed services, select all the Auto-Sync services that you had previously disabled and click "Enable". Once this is done, select them all and click "Start".

12. Click on one of the service names and click "Run Now" to start the sync. On the services screen, the bar that displays "Online" should change to "Task 1 of 1; 0.00 KB of x.xx". This may display as "Task 0 of 0" for a short while but should change quite quickly to Task 1 of 1. If this does not change there is still a problem with the replication. At this point a call will need to be logged with your third-party support.